The increasing prevalence of obesity in U.S. children and adolescents is a major health threat to our society, especially among minority and low social economic status (SES) populations. During adolescence physical activity (PA) decreases and is likely an important contributor to this alarming trend in childhood obesity rates. Although school-based behavioral interventions to address PA in children and adolescents have been conducted, little evidence suggests that these curriculum-based approaches lead to increases in overall PA. The present proposal will evaluate the efficacy of an innovative student-centered after-school program for promoting lifestyle changes in PA (7-day accelerometer assessment) among underserved adolescents (e.g., minorities, low SES). The student-centered intervention is consistent with Self-Determination (Motivation) Theory and Social Cognitive Theory in that it emphasizes increasing intrinsic motivation and behavioral skills for PA. Adolescents in the student-centered program take ownership in developing the program, select activities that generate fun and interest, and participate in generating their own coping strategies for making effective PA behavior changes during a videotaped session. Preliminary data is presented from our research team that demonstrates the feasibility of the student-centered PA program on increasing moderate-to-vigorous PA (MVPA) in underserved adolescents in South Carolina. The present application will use a school-based nested cohort design to evaluate the efficacy of a 6-month student-centered program versus a typical after school program on increasing PA at post-intervention and at 6-months follow-up in underserved adolescents. A total of 24 middle schools (120 6 th graders per school; N=2,880), located in South Carolina will be randomly assigned to one of two after-school programs (student-centered intervention, typical after-school program). We will analyze the data using a repeated measures analysis of covariance using SAS/PROC MIXED methods as outlined by Murray. In addition, mediational analyses will examine the association between psychosocial variables (PA self-efficacy, self-concept, motivation, social support, and enjoyment) and MVPA. Mediation will be tested using regression and structural equation modeling techniques as outlined by Baron and Kenny. This study will address an important public health problem that has implications for decreasing obesity in children and adolescents. [unreadable] [unreadable]